Intel Embree Delivers Massive Boost In Ray Tracing Performance For Arc GPUs

Intel Embree Delivers Massive Boost In Ray Tracing Performance For Arc GPUs 1

Intel has finally released a revamped version of its highly-popular Intel Embree Ray Tracing library, elevating the ray-tracing experience on the Arc GPUs to new heights.

Intel Embree 4.2 Version Supercharges Ray Tracing Performance On Arc GPUs, Surpasses NVIDIA's Counterpart

To recap, Intel Embree is an "open source library of ray tracing kernels" utilized by professionals in the media industry to enhance the performance of rendering platforms and speed up the process. As highlighted by Intel, the feature has received the " Scientific & Technical Achievement Award" for its contributions to the industry for various companies like Chaos V-Ray, Dreamworks MoonRay, Mercenaries Engineering Guerilla, Maxon Cinema 4D, and many more.

With the need for innovation in the media industry, the influence of Intel's Embree technology is growing immensely, which is why the new Intel Embree 4.2 version has set up a higher threshold. The significant change brought into the latest release is the integration of oneAPI’s SYCL implementation for Intel's discrete GPU (Arc lineup). For those who are unfamiliar with SYCL, it is similar to a single-architecture proprietary language, providing developers greater control over their respective code, which includes reusing across hardware targets and tuning it for a specific accelerator.

The SYCL is one of the newer Intel's oneAPI toolkits, which has also expanded to the Arc lineup. The Intel Embree 4.2 has granted developers the ability to "write C++ code for either CPUs or GPUs independently, " reducing developing time and code maintenance. Intel has benchmarked its Arc A750 against the i9-12900K CPU to demonstrate the feature's capabilities, both on a single source renderer.

The above results show that the Arc platform will significantly boost after this update, especially in commercial applications. Intel has provided consumers with the option to switch to Arc GPUs for faster rendering performance which is not only a cost-effective move but, in some scenarios, much superior to the rendering performance provided by a CPU.

Comparing a feature without its competitors doesn't complete the whole point. Judging by this, Intel has also benchmarked the Arc A750 against its nearest NVIDIA competitor, the GeForce RTX 3060. NVIDIA utilises the NVIDIA OptiX and Vulkan features to enhance rendering performance hence when the two technologies are compared, the results are interesting. Intel benchmarked both GPUs using the publicly available ChameleonRT path tracer.

The Intel Arc A750 surpasses NVIDIA's GeForce RTX 3060 at OptiX vs Embree comparison whilst showing similar results with Vulkan. Intel has indeed closed the gap regarding upscaling and rendering technologies.

Intel Embree 4.2 together with Intel GPUs is another step forward in our commitments to graphics acceleration and promoting an open software stack and ecosystem. We make our technologies broadly available and scalable to bring advanced ray tracing across laptops, workstations, data center/render farm, cloud—and to the world’s largest supercomputers.

-Intel

Moreover, Intel Embree 4.2 will also come with backward compatibility to support prior API features. The technology has also gained support in Blender's 3.6 LTS, and Intel has also provided a demonstration to showcase the capabilities of Intel's Embree on Blender.

Intel's Embree 4.2 is yet to be released in oneAPI Rendering Toolkit; however, you can find the open-source library on GitHub. It is exciting to see what Intel has in stores for consumers, especially since the company has been active in elevating the Arc platform, enhancing its capabilities for both gamers and professionals. The library is available to download here & developers can also take advantage of SYCL which includes easy migration from CUDA source codes so that it can be deployed across various devices (Architecture/Vendor Agnostic).

Written by Muhammad Zuhair

Post a Comment

0 Comments